PURPOSE: Atherosclerosis of peripheral cerebral arteries can lead to stroke either by stenosis formation or plaque rupture. This pathology is initiated by the alteration of arterial wall mechanical properties shown to be assessable by ultrasound elastography. Recently, noninvasive vascular elastography (NIVE) was introduced for noninvasive imaging of the mechanical properties of superficial arteries as markers of vulnerable plaques. However, NIVE motion estimates are angle-dependent, with optimal scanning angle being represented by the alignment of tissue motion with ultrasound beam orientation. The objective of this study was to introduce a model that compensates for such angle-dependence in order to reduce the bias on strain estimates, namely, when investigating longitudinal vessel segments. METHODS: The model is based on the Lagrangian speckle model estimator (LSME) because the LSME assesses the 2D-deformation matrix required to compute the scanning angle. RESULTS: Experiments on vessel-mimicking phantoms indicated that such a model enables the estimation of scanning angle with less than 3-degrees error. The method was also validated in vivo in human carotid arteries where less than 4-degrees error was observed. In both cases, the compensative model estimated the inclination angles with low variability. CONCLUSION: Angle-dependence may be an important factor to consider in avoiding potentially distort clinical diagnoses. Results, reported in this article, suggest that the LSME-based compensative model might be considered as a very interesting and promising clinical tool for NIVE applications.
PURPOSE:Atherosclerosis of peripheral cerebral arteries can lead to stroke either by stenosis formation or plaque rupture. This pathology is initiated by the alteration of arterial wall mechanical properties shown to be assessable by ultrasound elastography. Recently, noninvasive vascular elastography (NIVE) was introduced for noninvasive imaging of the mechanical properties of superficial arteries as markers of vulnerable plaques. However, NIVE motion estimates are angle-dependent, with optimal scanning angle being represented by the alignment of tissue motion with ultrasound beam orientation. The objective of this study was to introduce a model that compensates for such angle-dependence in order to reduce the bias on strain estimates, namely, when investigating longitudinal vessel segments. METHODS: The model is based on the Lagrangian speckle model estimator (LSME) because the LSME assesses the 2D-deformation matrix required to compute the scanning angle. RESULTS: Experiments on vessel-mimicking phantoms indicated that such a model enables the estimation of scanning angle with less than 3-degrees error. The method was also validated in vivo in human carotid arteries where less than 4-degrees error was observed. In both cases, the compensative model estimated the inclination angles with low variability. CONCLUSION: Angle-dependence may be an important factor to consider in avoiding potentially distort clinical diagnoses. Results, reported in this article, suggest that the LSME-based compensative model might be considered as a very interesting and promising clinical tool for NIVE applications.